Glycine max seeds are normally inoculated with Bradyrhizobium japonicum before planting, but if I am attempting to see differences in nodulation between treatments with addition of a potentially toxic material I would rather inoculate the soil in pot experiments, because the bacterium is not native in my area and thus would be unlikely to find. The point is to draw a parallel with areas that has previous history of G. max growth and would not necessarily need seed inoculation because B. japonicum would already be present in the soil. Or would it be irrelevant to think this way, because a lot of farmers inoculates every growth season?